In a recent book, Donald Norman describes how products developed from lists of user demands often fail when they are integrated into work environments. According to Norman, failures occur because actual customers for a product and their needs are often different from the focus groups that were interviewed during product development. Further, Norman states that the design process is often a linear sequence of operations that inhibits interactions among members of the team. He recommends that designs should emerge from a process of observing customers at work together with frequent prototype evaluations by users and the design team including managers, engineers, software specialists, technical writers, and behavioral scientists.

The Air Force Research Laboratory, Warfighter Training Research Division, has applied this human-centered development process to create a four-ship, F-16 simulation testbed. The testbed has been used in a series of Distributed Mission Training (DMT) exercises with pilots and air weapons controllers. The goal of these exercises has been to identify and document how DMT systems can be designed and used most effectively to enhance mission skills. The lab's design team has used these exercises to observe warfighters at work and identify significant training needs. During and after each training exercise, feedback is used to determine how to refine the scenarios, procedures, and testbed systems to support a training environment that complements current flying training requirements. Exercises have included RoadRunner 98 (a composite force exercise), an air-to-air training evaluation study, and flight lead upgrade training. Collected data are being used to determine DMT strengths and weaknesses and to identify the best training uses of DMT. These data are also providing a basis for a performance measurement system to assess the effectiveness of training in DMT.